Workshop

Ship an AI Agent to Production on AWS in One Day

Saurav Sharma

AI cost consultant, 6 years at Amazon

A public AI agent endpoint in your AWS account, with a hard cost limit

One live Sunday, 6 hours. You deploy an AI agent to your own AWS account with Bedrock and AgentCore. Then you add the production parts: a Lambda tool behind API Gateway, a guardrail, WAF, and a hard cost limit.

At 4:45 pm ET you have a public HTTPS agent endpoint in your account, a guardrail that stops 10 attack prompts, a budget alarm, and a teardown command.

I spent 6 years at Amazon, first with enterprise customers on cloud architecture and cost, then as an engineer on LLM and GenAI platforms. I hold 12 AWS certifications. I also paid for my own mistake: Bedrock calls cost me $47.53 before I added a gate. Hour 5 gives you that gate.

You get the recording and the repo. Each class has a limit of 20 seats, because AWS account setup needs support.

Founding cohort rate: $399 instead of $549. These are the first dates, so the price is lower. In return, I ask for your honest feedback after the day.

Join once, come back free: you can join any later date at no cost, when a seat is open. Email me and I add you.

What you’ll learn

At 4:45 pm ET you have a public HTTPS agent endpoint in your account, a guardrail, a budget alarm, and a teardown command.

  • Call a model from code under a least-privilege IAM role, with the Converse API.

  • Deploy a small agent to AgentCore Runtime, with session memory.

  • Add one Lambda tool that the agent calls.

  • Put API Gateway with auth in front, so you have a public HTTPS endpoint.

  • Set a Bedrock Guardrail for denied topics, PII, and prompt attacks.

  • Add a WAF rate rule, and run a sheet of 10 attack prompts against the endpoint.

  • Log tokens per call, find your cost per 1,000 calls, and set a Budget alarm with an action.

  • Leave with traces in CloudWatch, a 1-page security review sheet, and one teardown command.

Workshop agenda

  • 10:00 am ET: Bedrock in your account

    Model access, a least-privilege IAM role, and the Converse API. You leave with a working model call from code, under a role you can show a reviewer.

  • 11:00 am: The agent on AgentCore Runtime

    You leave with a small agent deployed to AgentCore, with session memory.

  • 12:00 pm: Lunch

    A 45-minute break.

  • 12:45 pm: Tools and the front door

    One Lambda tool, and API Gateway with auth. You leave with a public HTTPS endpoint, and the agent calls your Lambda.

  • 1:45 pm: Guardrails and WAF

    A Bedrock Guardrail (denied topics, PII, prompt attack) and a WAF rate rule. You leave with a 10-prompt attack sheet that your guardrail stops.

  • 2:45 pm: Cost limits

    Tokens logged per call, your cost per 1,000 calls, and a Budget alarm with an action. I show the $47.53 bill that made me add a gate.

  • 3:45 pm: Production checklist and teardown

    Traces in CloudWatch, a 1-page security review sheet, and a one-command teardown. We end at 4:45 pm.

Learn directly from Saurav

Saurav Sharma

Saurav Sharma

Contact

I teach from my own code, bills, and mistakes. 6 years at Amazon.

See all products from Saurav

Who this workshop is for

  • Cloud and backend engineers who know AWS basics and want an agent in production, with security and cost limits.

  • Tech leads and architects who must show a reviewer how the agent is secured and what each call costs.

  • This workshop does not fit people with no AWS account or no command-line practice.

Prerequisites

  • Your own AWS account with admin rights

    Use a personal account. A work account with SCP limits can block the labs. Set an AWS Budget of $20 before class.

  • Bedrock model access in us-east-1

    Turn on model access and make one test call that returns text. I send a checklist 7 days before class.

  • AWS CLI v2 and Python 3.12

    Install both and pass the checklist by the Thursday before class.

What's included

Saurav Sharma

Live sessions

Learn directly from Saurav Sharma in a real-time, interactive format.

Lifetime access

Go back to course content and recordings whenever you need to.

Community of peers

Stay accountable and share insights with like-minded professionals.

Certificate of completion

Share your new skills with your employer or on LinkedIn.

Maven Guarantee

Your purchase is backed by the Maven Guarantee.

Frequently asked questions

Maven for Teams

Reimbursement

Get your company to pay

Everything L&D needs: email template, receipts, and certificate of completion.

Get reimbursed

Team discount

Learn with your teammates

Save 20%+ when 2 or more teammates enroll in the same cohort.

Save 20%+ with a team

Private cohort

Run a cohort for your org

A dedicated cohort with a custom schedule and curriculum, tailored to your team.

Book a private cohort

$399

USD

Nov 29
·

10am–5pm EST

·

4 cohorts